PetNourish.app
AI-Powered Pet Nutrition
This one started with my dog. Commercial pet food ingredient lists read like a chemistry experiment, and the homemade alternatives online are either nutritionally incomplete or authored by people who clearly don't understand canine dietary requirements. PetNourish.app generates custom homemade meal recipes that actually hit the right nutritional targets.
Pet nutrition has real stakes — the wrong calcium-to-phosphorus ratio in a growing puppy can cause skeletal problems. This isn't a space for confident AI hallucinations. The challenge was building a generation system that understood breed-specific requirements, life stage differences, and the genuinely toxic ingredient list (onions, grapes, xylitol — the list is longer than most people realise) well enough to produce recipes that were both palatable and safe.
Stack & Architecture
- AI recipe generation — custom-prompted for species, breed, age, and health conditions
- Nutritional database — AAFCO-aligned nutrient targets by life stage
- Ingredient safety layer — cross-references known toxic ingredients per species
- Cloudflare Workers — edge-deployed for fast, low-latency recipe generation
Nutritional Accuracy First
Every generated recipe is validated against AAFCO minimum nutrient profiles before it's returned to the user. The AI generates; the validator checks. If a recipe comes back short on zinc or runs hot on Vitamin A, it either regenerates or surfaces the gap explicitly. The recipe generation prompt is tightly constrained to nutritional targets, not just taste.
Breed and Life Stage Intelligence
A 6-month-old Great Dane puppy has fundamentally different calcium needs than a 3-year-old Beagle. Giant breed puppies are particularly sensitive to excess calcium during growth. The system knows this, and adjusts targets accordingly rather than applying a one-size-fits-all profile.
Ingredient Safety
Before any recipe is finalised, every ingredient passes through a species-specific toxicity check. The database covers known toxins across dogs, cats, and common exotic pets. It's not just the obvious ones — certain fish prepared certain ways carry risks most people don't know about.
My dog has been eating PetNourish-generated meals for over a year. His coat condition and energy levels are objectively better than on commercial food. The app is live and has a small but growing user base of owners who take their pet's nutrition as seriously as their own.
Want to build something like this?
Fractional CTO, technology advisory, or a build from scratch. Let's talk.